On the association between spring Arctic sea ice concentration

and Chinese summer rainfall

Bingyi Wu,1 Renhe Zhang,1 Bin Wang,2 and Rosanne D’Arrigo3

Received 14 January 2009; revised 26 February 2009; accepted 12 March 2009; published 5 May 2009.

[1] A statistical relationship between spring Arctic sea iceconcentration (SIC) and Chinese summer rainfall is identifiedusing singular value decomposition (SVD). Results showthat decreased (increased) spring SIC in the Arctic Ocean andthe Greenland Sea corresponds to increased (decreased)summer rainfall in northeast China and central Chinabetween the Yangtze River and the Yellow River (28�–36�N); and decreased (increased) rainfall in south China.Corresponding summer 500 hPa height anomalies show aEurasian wave train structure, which originates in northernEurope and extends southeastwards to northeast China, and asouth-north dipole structure over East Asia south to LakeBaikal. Such a spatial distribution of 500 hPa heightanomalies and corresponding summer rainfall anomaliescannot be entirely attributed to the impact of East Asiansummer monsoon (EASM) variability. The spring Arctic SICprovides a complementary precursor for Chinese summerrainfall variability. The combined impacts of both springArctic SIC and Eurasian snow cover on the Eurasian wavetrain may explain their statistical linkage. Citation: Wu, B.,

R. Zhang, B. Wang, and R. D’Arrigo (2009), On the association

between spring Arctic sea ice concentration and Chinese summer

rainfall, Geophys. Res. Lett., 36, L09501, doi:10.1029/

2009GL037299.

1. Introduction

[2] The role of Arctic sea ice in the Earth’s climatesystem has been extensively investigated using observationsand simulations [Walsh, 1983; Honda et al., 1996; Rigor etal., 2002; Alexander et al., 2004; Deser et al., 2004;Magnusdottir et al., 2004; Wu et al., 2004]. Yet, theassociation of Arctic sea ice with Chinese summer rainfalland East Asian summer monsoon (EASM) is still unclearand not well studied. Limited research [Niu et al., 2003] hasshown that reduced spring sea ice extent in the Bering andOkhotsk seas corresponds to enhanced June–July rainfall insoutheast China. However, spring sea ice variability isgenerally out of phase between the Bering and Okhotskseas, thus grouping them together is inappropriate [Cavalieriand Parkinson, 1987; Fang and Wallace, 1994]. The moti-vation of this study is to assess potential predictors forChinese summer rainfall variations. We therefore investigatea statistical linkage between Arctic sea ice variability in the

boreal spring (March–May) and Chinese summer rainfall.Because Arctic sea ice concentration (SIC) variations typi-cally have a more substantial impact on the atmosphericcirculation than sea ice extent variations [Alexander et al.,2004] we focus herein on spring Arctic SIC variability.

2. Data and Method

[3] The primary datasets used in this study are themonthly mean Arctic SIC dataset (on a 1� latitude � 1�longitude grid for the period 1961–2007), obtained fromthe British Atmospheric Data Centre (BADC, http://badc.nerc.ac.uk/data/hadisst/), and a monthly 513-station rainfalldataset for China, obtained from the National Meteorolog-ical Information Centre of China, spanning the period from1968 to 2005. The sea level pressure (SLP), 850 hPa winds,and 500 hPa geopotential heights are obtained from theNational Center for Environmental Prediction (NCEP) andthe National Center for Atmospheric Research (NCAR)reanalysis datasets for the period of 1968–2007. Themonthly mean snow water equivalent (SWE) dataset wasderived from National Snow and Ice Center during theperiod from 1979 to 2004 [Armstrong et al., 2005]. TheSWE data have been converted to 1� � 1� grids in thisstudy.[4] To reveal the dominant modes of co-variability be-

tween spring Arctic SIC and Chinese summer rainfall, weused the singular value decomposition (SVD) method.Additionally, we applied empirical orthogonal function(EOF) analysis (based on the correlation matrix) to extract,respectively, dominant modes of spring Arctic SIC variabil-ity and Chinese summer rainfall variability. In this study, wefocus on interannual variability of the seasonal means,averaged for three spring (March–May) and three summer(June–August) months.

3. A Statistical Linkage Between Spring ArcticSIC and Chinese Summer Rainfall

[5] Figure 1 shows the time series and correspondingspatial distributions of the leading SVD between springArctic SIC and China summer rainfall, which accounts for19% of their co-variance. Both Arctic SIC and summerrainfall in the leading SVD display a coherent interannualvariability (the correlation is 0.83, Figure 1a) and apparentinterdecadal variations with turning points occurring around1978 and 1992, respectively. For spring SIC, negativephases were frequent during the periods 1968–1978 and1993–2005, and separated by dominant positive phases.The corresponding SIC anomalies were positive in most ofEurasian marginal seas, with negative SIC anomalies in theArctic Basin, the Beaufort Sea, and the Greenland Sea

GEOPHYSICAL RESEARCH LETTERS, VOL. 36, L09501, doi:10.1029/2009GL037299, 2009ClickHere

for

FullArticle

1Chinese Academy of Meteorological Sciences, China MeteorologicalAdministration, Beijing, China.

2Department of Meteorology and International Pacific Research Center,University of Hawaii at Manoa, Honolulu, Hawaii, USA.

3Tree-Ring Laboratory, Lamont-Doherty Earth Observatory, Palisades,New York, USA.

Copyright 2009 by the American Geophysical Union.0094-8276/09/2009GL037299$05.00

L09501 1 of 6

during 1979–1992. Opposite SIC anomalies occurred dur-ing the periods 1968–1978 and 1993–2005 (Figure 1b). Itis worth noting that spring SIC variability is out of phase inthe Bering Sea and the Okhotsk Sea. During the period1979–1992, positive rainfall anomalies frequently appearedin northeast China and central China between the YangtzeRiver and the Yellow River (28�–36�N), with negativerainfall anomalies in south and southeast China (Figure 1c).Opposing spatial distribution of summer rainfall anomaliesfrequently occurred in the period 1968–1978 and 1993–2005. Ding et al. [2007] examined interdecadal variations ofChinese summer rainfall over the past five decades andfound that 1978 and 1992 are two abrupt climate changepoints, which is well consistent with our results.[6] The heterogeneous correlations for the leading SVD

show significant negative correlations in the Arctic Ocean,the Greenland Sea (Figure 2a). For summer rainfall, signif-icant positive correlations are seen in northeast China and

central China between the Yangtze and the Yellow Rivers,with significant negative correlations in south and southeastChina (Figure 2b). These results imply that decreased springSIC in the Arctic Ocean and the Greenland Sea tends to be

Figure 1. (a) Normalized time series of spring Arctic SIC(solid line) and summer rainfall (dashed line) variations inthe leading SVD, the red and green lines denote their 5-yearrunning means, respectively, (b) the spatial distribution ofspring Arctic SIC variations in the leading SVD, (c) sameas Figure 1b but for the summer rainfall variations. InFigures 1b and 1c units are arbitrary.

Figure 2. (a) The spatial distribution of spring Arctic SICheterogeneous correlations, the shading areas denotecorrelations at the 0.05 and 0.01 significance levels, respec-tively, (b) same as in Figure 2a but for Chinese summerrainfall, (c) the spatial distribution of the leading EOF ofChinese summer rainfall (Units: arbitrary), (d) normalizedtime series of both the leading EOF of summer rainfall (solidline) and summer rainfall variations in the leading SVD(dashed line), the red and green lines denote their 5-yearrunning means, respectively.

L09501 WU ET AL.: SEA ICE AND CHINESE SUMMER RAINFALL L09501

2 of 6

associated with increased summer rainfall in northeastChina and central China between the Yangtze and theYellow Rivers and decreased summer rainfall in south andsoutheast China.[7] We applied EOF analysis to spring SIC during the

period 1961–2007. The first two EOFs account for 13%and 10% of the variance, respectively. According to Northet al. [1982], the first three EOFs are well separated fromeach other. In fact, the leading EOF reflects the overalldecline trend in sea ice associated with global warming (notshown). The second EOF of spring Arctic SIC variabilityclosely resembles Figure 1b but with opposite sign (notshown). The correlation between the second EOF andspring SIC variations in the leading SVD is �0.95. Thusspring SIC variations in the leading SVD reflect a dominantpattern of spring SIC variability.[8] Similarly, we extracted the leading EOF of Chinese

summer rainfall variability, which accounts for 10% of thevariance during the period 1968–2005. The spatial distri-bution and temporal evolution of the leading rainfall EOFclosely resemble Figures 1c and 1a, respectively, but withopposite sign (Figures 2c and 2d). The correlation betweenthem is �0.81. On interdecadal time-scales, negative phasesare dominant during the period 1979–1991, in contrast toopposite phases before and after (Figure 2d). Consequently,summer rainfall variation in the leading SVD is significantlycorrelated with the leading rainfall EOF.[9] In fact, the spatial distribution of the leading rainfall

EOF can be attributed to the combination effect of two wavetrains in summer 500 hPa height anomalies (Figure 3a). Thefirst wave train shows a meridional structure along the EastAsian coast, with two negative anomalous centers oversoutheast China and the Okhotsk Sea and a positiveanomalous center between them when the rainfall patternis in its positive phase. The wave train closely resembles thePacific-Japan pattern [Nitta, 1987; Huang and Li, 1987],which is closely related to EASM variability [Wang et al.,2001; Wu et al., 2008]. The second wave train originatesfrom northern Europe extending southeastwards to KoreanPeninsula. There are three anomalous centers over northernEurope, northwestern Asian continent and northeast Asia.Correspondingly, summer 850 hPa wind anomalies show ananomalous cyclone over south and southeast China and ananomalous anticyclone over northeast Asia (Figure 3b).Figure 3b shows opposing expressions of EASM variabilityin north China and south and southwest China. There is ananomalous anticyclone over northeast China, leading todecreased summer rainfall there (Figure 3c). Divergencesin anomalous 850 hPa winds seem to appear over theYangtze River valley, with an anomalous cyclone over37.5�–50�N, 90�–110�E. Such a spatial distribution ofsummer 850 hPa wind anomalies is dynamically consistentwith significantly increased rainfall in the mid-reaches ofthe Yellow River valley, southwest, south and southeastChina; with significantly decreased rainfall in northeastChina, the mid-low reaches of Yangtze River (Figure 3c).[10] Figure 3d shows the 500 hPa height anomalies

associated with the circulation index of the western NorthPacific (WNP) and EASM proposed by Wang and Fan[1999]. The Wang-Fan index describes the precipitationvariability in the WNP monsoon trough, which drivessubtropical and extratropical EASM variability through

the meridional tri-pole teleconnection (Figure 3d). TheWang-Fan index captures the tri-pole wave train very well,however, it cannot completely capture the Eurasian wavetrain shown in Figure 3a, and two anomalous centers oversouth of 50�Nmigrate eastwards relative to that in Figure 3a.Corresponding 850 hPa wind anomalies obviously differfrom that in Figure 3b, particularly over area from northChina to Lake Baikal where wind anomalies are very weak,leading to significantly decreased rainfall along the YangtzeRiver valley, the low-reaches of the Yellow River andnortheast China and moderate increased rainfall in parts ofsouth and southeast China (Figure 3f). Although the leadingrainfall EOF is significantly correlated with the Wang-Fanindex (r = 0.43, at the 0.01 significance level) andcorresponding height anomalies show the great similarityalong the East Asian coast region WNP-EASM cannotcompletely explain the spatial distribution of summer rain-fall anomalies associated with the leading rainfall EOF,particularly in north China.[11] Summer 500 hPa height anomalies associated with

the preceding spring Arctic SIC also show the Eurasianwave train originating from northern Europe and extendingsoutheastwards to southeast of Lake Baikal region andnortheast China (Figure 3g). Height anomalies exhibit asouth-north dipole structure south of Lake Baikal. Ananomalous cyclone over northeast China and an anomalousanticyclone over south China are concurrent, with strongconvergences over the Yangtze River valley (Figure 3h).Thus, significantly increased rainfall can be observed innortheast China and central China between the YangtzeRiver and the Yellow River, and significantly decreasedrainfall in south and southeast China (Figure 3i). Since theWang-Fan index reflects WNP-EASM variability that isdominated by ENSO, the results here suggested that thespring Arctic SIC provides a precursor for Chinese summerrainfall variability that is complementary to the effects ofENSO. Although spring Arctic SIC and Chinese summerrainfall exhibit coherent interannual and interdecadal varia-tions their association on interannual time scales is moreimportant relative to interdecadal time scales (not shown).

4. Possible Linkage Mechanisms and Discussion

[12] We only attempt to explain the possible reason forthe Eurasian wave train. In fact, variations in spring ArcticSIC and Eurasian snow cover are coherent, which isreflected in significant positive correlations between springArctic SIC in the leading SVD and the gridded SWE innorthern Eurasia (not shown). This result implies thatdecreased Arctic SIC in the Arctic Ocean and the GreenlandSea and increased SIC in most of Eurasian marginal seas(Figure 1b) correspond to excessive spring Eurasian SWE.The spring regionally-averaged SWE (55�–75�N and 60�–120�E) is broadly similar to Arctic SIC variations in theleading SVD (r = 0.64, at the 0.01 significance level), butindividual springs can be radically different (Figure 4a). Theevolution of 500 hPa height anomalies from spring tosummer associated with spring Eurasian SWE is illustratedin Figures 4b and 4c, which show composite differencesbetween above (s > 0.8: 1979, 1980, 1983, 1984, 1985,1987, 1996) and below-normal Eurasian SWE (s < �0.8:1990, 1995, 1997, 2001, 2002). Positive anomalies in spring

L09501 WU ET AL.: SEA ICE AND CHINESE SUMMER RAINFALL L09501

3 of 6

Eurasian SWE correspond to three anomalous centers,respectively, over the western and northern Europe, andnorthwest of Lake Baikal (Figure 4b). The three anomalouscenters then migrate, respectively, to the northern Europe,northwest Asia continent close to the Kara Sea, and south-east of Lake Baikal (Figure 4c). Figure 4c closely resemblesFigure 3g but with strengthened amplitudes, reflecting thecombined effect of both spring Arctic SIC and EurasianSWE [Yasunari et al., 1991]. The simulation result partly

supports this hypothesis [Dethloff et al., 2006]. It should bepointed out that differences in analysis approach and datalength are likely to induce the difference in amplitude inFigures 3g and 4c.[13] It is very difficult, however, to directly detect the

impact of sea ice on the atmosphere because of dominantroles of atmospheric forcing on sea ice [Deser et al., 2000;Wu et al., 2004]. We speculate that surface wind forcingdetermines spring SIC anomalies, which largely persist until

Figure 3. Summer anomalies in (a) 500 hPa height, (b) 850 hPa wind, and (c) Chinese summer rainfall, derived fromregressions on the leading EOF of Chinese summer rainfall variability, in Figures 3a and 3b, the shading areas representheight and wind velocity anomalies exceed the 0.05 and 0.01 confidence levels, respectively, in Figure 3c, thin and thicklines denote that rainfall anomalies are above the 0.05 and 0.01 confidence levels, respectively. Units: in Figure 3c, gpm; inFigure 3b, m/s; in Figure 3c, 0.1 mm. (d–f and g–i) same as Figures 3a–3c, respectively, but derived from the circulationindex of Wang and Fan [1999] and spring Arctic SIC signs in the leading SVD.

L09501 WU ET AL.: SEA ICE AND CHINESE SUMMER RAINFALL L09501

4 of 6

the following summer. In turn, the accumulated impact ofpersistent SIC anomalies from spring to summer couldimpact the summer subpolar westerlies through changingalbedo and heat flux between the atmosphere and ocean,which further influences redistribution of Rossby waves(Figure 3g). The precise mechanism responsible for theirassociations needs to be further investigated through obser-vations and simulations in the future.

[14] Acknowledgments. This study was supported by the NationalKey Basic Research and Development Project of China (grants2004CB418300 and 2007CB411505), Chinese COPES project(GYHY200706005) and the National Natural Science Foundation of China(grant 40875052).

ReferencesAlexander, M. A., et al. (2004), The atmospheric response to realistic seaice anomalies in an AGCM during winter, J. Clim., 17, 890–905.

Armstrong, R. L., M. J. Brodzik, K. Knowles, and M. Savoie (2005), GlobalMonthly EASE-Grid Snow Water Equivalent Climatology, http://nsidc.org/data/nsidc-0271.html, Natl. Snow and Ice Data Cent., Boulder, Colo.

Cavalieri, D. J., and C. L. Parkinson (1987), On the relationship betweenatmospheric circulation and fluctuations in sea ice extents of the Beringand Okhotsk Seas, J. Geophys. Res., 92, 7141–7162.

Dethloff, K., et al. (2006), A dynamical link between the Arctic and theglobal climate system, Geophys. Res. Lett., 33, L03703, doi:10.1029/2005GL025245.

Deser, C., J. E. Walsh, and M. S. Timlin (2000), Arctic sea ice variability inthe context of recent atmospheric circulation trends, J. Clim., 13, 617–633.

Deser, C., G. Magnusdottir, R. Saravanan, and A. Phillips (2004), Theeffects of North Atlantic SST and sea ice anomalies on the winter circula-tion in CCM3. Part II: Direct and indirect components of the response,J. Clim., 17, 877–889.

Ding, Y., Z. Wang, and Y. Sun (2007), Inter-decadal variation of the sum-mer precipitation in East China and its association with decreased Asiansummer monsoon. Part I: Observational evidences, Int. J. Climatol., 28,1139–1161, doi:10.1002/joc.1615.

Fang, Z., and J. M. Wallace (1994), Arctic sea ice variability on a timescaleof weeks: Its relation to atmospheric forcing, J. Clim., 7, 1897–1913.

Honda, M., K. Yamazaki, Y. Tachibana, and K. Takeuchi (1996), Influenceof Okhotsk sea-ice extent on atmospheric circulation, Geophys. Res. Lett.,23, 3595–3598.

Huang, R., and W. Li (1987), Influence of the heat source anomaly over thewestern tropical Pacific on the subtropical high over east Asia, in Pro-ceedings of the International Conference on the General Circulation ofEast Asia, pp. 40–51, Chengdu, China.

Magnusdottir, G., C. Deser, and R. Saravanan (2004), The effects of NorthAtlantic SST and sea ice anomalies on the winter circulation in CCM3.Part I: Main features and storm track characteristics of the response,J. Clim., 17, 857–876.

Nitta, T. (1987), Convective activities in the tropical western Pacific andtheir impacts on the NorthernHemisphere summer circulation, J.Meteorol.Soc. Jpn., 65, 165–171.

Niu, T., P. Zhao, and L. Chen (2003), Effects of the sea-ice along the NorthPacific on summer rainfall in China, Acta Meteorol. Sin., 17, 52–64.

North, G. R., T. L. Bell, R. F. Cahalan, and F. J. Moeng (1982), Samplingerrors in the estimation of EOFs, Mon. Weather Rev., 110, 699–706.

Rigor, I. G., J. M. Wallace, and R. L. Colony (2002), Response of sea ice tothe Arctic Oscillation, J. Clim., 15, 2648–2663.

Walsh, J. E. (1983), Role of sea ice in climate variability: Theories andevidence, Atmos. Ocean, 21, 229–242.

Wang, B., and Z. Fan (1999), Choice of south Asian summer monsoonindices, Bull. Am. Meteorol. Soc., 80, 629–638.

Figure 4. (a) Normalized time series of spring regionally averaged snow water equivalent (55�–75�N and 60�–120�E)(purple line) and spring Arctic SIC in the leading SVD, (b) composite differences in spring mean 500 hPa heights betweenexcessive and less snow water equivalent (for composite cases, see the text), the shading areas denote that differencesexceed the 0.05 and 0.01 confidence levels, respectively, (c) same as Figure 4b but for summer. In Figures 4b and 4c, unitsare gpm.

L09501 WU ET AL.: SEA ICE AND CHINESE SUMMER RAINFALL L09501

5 of 6

Wang, B., R. Wu, and K. Lau (2001), Interannual variability of the Asiansummer monsoon: Contrasts between the Indian and the western NorthPacific–east Asian monsoon, J. Clim., 14, 4073–4090.

Wu, B., J. Wang, and J. Walsh (2004), Possible feedback of winter sea icein the Greenland and the Barents Sea on the local atmosphere, Mon.Weather Rev., 132, 1876–1968.

Wu, B., R. Zhang, Y. Ding, and R. D’Arrigo (2008), Distinct modes of theeast Asian summer monsoon, J. Clim., 21, 1122–1138.

Yasunari, T., A. Kitoh, and T. Tokioka (1991), Local and remote responsesto excessive snow mass over Eurasia appearing in the northern spring and

summer climate: A study with the MRI-GCM, J. Meteorol. Soc. Jpn., 69,473–487.

�����������������������R. D’Arrigo, Tree-Ring Laboratory, Lamont-Doherty Earth Observatory,

61 Route 9W, Palisades, NY 10964, USA.B. Wang, Department of Meteorology and International Pacific Research

Center, University of Hawaii at Manoa, 1680 East West Road, PostBuilding 401, Honolulu, HI 96822-2219, USA.B. Wu and R. Zhang, Chinese Academy of Meteorological Sciences,

China Meteorological Administration, No. 46 Zhong-Guan-Cun SouthAvenue, Beijing 100081, China. (wby@cams.cma.gov.cn)

L09501 WU ET AL.: SEA ICE AND CHINESE SUMMER RAINFALL L09501

6 of 6